Critical length in electromigration-experiments and theory

New experiments to determine electromigration drift velocity were badly needed in the 1970’s. One such experiment employed a series of aluminum islands, or strips, deposited on top of non-electromigrating but electrically conducting TiN. The experiments showed that the edge displacement near the cathode of the strips drifted as a result of electrical current passage. Figure 1 shows both a plan view as well as a graph of this displacement (1). The displacement is clearly seen to be a function of strip length, namely, longer strips showed larger edge displacement. Very short stripes, such as the one on the right in figure 1, did not show any signs of electromigration. It was this kind of experiments that led to the suggestion that there is a back flow of atoms counteracting the electromigration. The back flow originates from a free energy gradient (for example, a stress gradient) set up by atom crowding near the anode. The back flow was assumed to be stronger in short stripes since the energy gradient drivi...